Packing materials for liquid chromatography (LC) are generally classified into two types: organic materials, e.g., polydivinylbenzene, and inorganic materials typified by silica. Many organic materials are chemically stable against strongly alkaline and strongly acidic mobile phases, allowing flexibility in the choice of mobile phase pH. However, organic chromatographic materials generally result in columns with low efficiency, leading to inadequate separation performance, particularly with low molecular-weight analytes. Furthermore, many organic chromatographic materials shrink and swell when the composition of the mobile phase is changed. In addition, most organic chromatographic materials do not have the mechanical strength of typical chromatographic silicas.
Due in large part to these limitations, silica is the material most widely used in High Performance Liquid Chromatography (HPLC). The most common applications employ silica that has been surface-derivatized with an organic group such as octadecyl (C18), octyl (C8), phenyl, amino, cyano, etc. As stationary phases for HPLC, these packing materials result in columns that have high efficiency and do not show evidence of shrinking or swelling.
Silica is characterized by the presence of silanol groups on its surface, During a typical derivatization process such as reaction with octadecyldimethylchlorosilane, at least 50% of the surface silanol groups remains unreacted. These residual silanol groups interact with basic and acidic analytes via ion exchange, hydrogen bonding and dipole/dipole mechanisms. The residual silanol groups create problems including increased retention, excessive peak tailing and irreversible adsorption of some analytes. Another drawback with silica-based columns is their limited hydrolytic stability. First, the incomplete derivatization of the silica leaves patches of bare silica surface which can be readily dissolved under alkaline conditions, generally pH>8.0, leading to the subsequent collapse of the chromatographic bed. Second, the bonded phase can be stripped off the surface under acidic conditions, generally pH<2.0, and eluted off the column by the mobile phase, causing loss of analyte retention, and an increase in the concentration of surface silanol groups.
To overcome the problems of residual silanol group activity and hydrolytic instability of silica-based stationary phases, many methods have been tried including use of ultrapure silica, carbonized silica, coating of the silica surface with polymeric materials, endcapping free silanol groups with a short-chain reagent such as trimethylsilane, and the addition of suppressors such as amines to the eluant. These approaches have not proven to be completely satisfactory in practice.
One approach is disclosed in U.S. Pat. No. 4,017,528. A process for preparing a hybrid silica is described wherein an alkyl functionality is coupled into both the skeleton structure and the surface of the silica. According to the '528 patent, the hybrid silica can be prepared by two methods. In the first method, a mixture of tetraethoxysilane (TEOS) and an organotriethoxysilane, e.g., alkyltriethoxysilane, is co-hydrolyzed in the presence of an acid catalyst to form a liquid material containing polyorganoethoxysiloxane (POS) oligomers, e.g., polyalkylethoxysiloxane oligomers. Then, the POS is suspended in an aqueous medium and gelled into porous hybrid particles in the presence of a base catalyst. In the second method, the material is prepared by a similar procedure except that the suspension droplet is a mixture of organotriethoxysilane, e.g., alkyltriethoxysilane, and polyethoxysiloxane (PES) oligomers; the latter is prepared by partial hydrolysis of TEOS.
There are several problems associated with the '528 hybrid material. First, these hybrid materials contain numerous micropores, i.e., pores having a diameter below about 34 Å. It is known that such micropores inhibit solute mass transfer, resulting in poor peak shape and band broadening.
Second, the pore structure of the '528 hybrid material is formed because of the presence of ethanol (a side product of the gelation process) within the suspension oil droplets. The pore volume is controlled by the molecular weight of the POS or PES. The lower the molecular weight of the POS or PES, the more ethanol is generated during the gelation reaction, and subsequently a larger pore volume is produced. However, part of the ethanol generated during the gelation is able to diffuse into the aqueous phase by partition. If the amount of the ethanol generated within the suspension droplets is too great, the partition of the ethanol will cause the structure of the droplets to collapse, forming irregularly-shaped particles as opposed to spherical particles. Therefore, the strategy to control the pore volume of the hybrid material described in the '528 patent has certain limitations, particularly for preparing highly spherical hybrid materials with a pore volume greater than about 0.8 cm3/g. It is well known in the art that irregularly-shaped materials are generally more difficult to pack than spherical materials. It is also known that columns packed with irregularly-shaped materials generally exhibit poorer packed bed stability than spherical materials of the same size.
Third, the '528 hybrid materials are characterized by an inhomogeneous particle morphology, which contributes to undesirable chromatographic properties, including poor mass transfer properties for solute molecules. This is a consequence of the gelation mechanism, where the base catalyst reacts rapidly near the surface of the POS droplet, forming a “skinned” layer having very small pores. Further gelation in the interior of the droplet is then limited by the diffusion of catalyst through this outer layer towards the droplet center, leading to particles having skeletal morphologies and hence pore geometries, e.g., “shell shaped”, which can vary as a function of location between the particle center and outer layer.
A further problem associated with silica particles and polymer particles is packed bed stability. Chromatography columns packed with spherical particles can be considered to be random close packed lattices in which the interstices between the particles form a continuous network from the column inlet to the column outlet. This network forms the interstitial volume of the packed bed which acts as a conduit for fluid to flow through the packed column. In order to achieve maximum packed bed stability, the particles must be tightly packed, and hence, the interstitial volume is limited in the column. As a result, such tightly packed columns afford high column backpressures which are not desirable. Moreover, bed stability problems for these chromatography columns are still typically observed, because of particle rearrangements.
In an attempt to overcome the problem of packed bed stability, several groups have reported studies on stabilizing the packed bed by sintering or interconnecting inorganic, e.g., silica based particles. In the sintering process, particles are joined to one another by grain boundaries. In one approach, previously prepared octadecylsilica particles are immobilized in a sol-gel matrix or a polymer matrix prepared in situ in a chromatography column. In another approach, agglomeration of the silica based C-18 particles at high temperature has been reported (M. T. Dulay, R. P. Kulkarni, R. N. Zare, Anal. Chem., 70 (1998) 5103; Xin, B.; Lee, M. L. Electrophoresis 1999, 20, 67; Q. Tang, B. Xin, M. L. Lee, J. Chromatogr. A, 837 (1999) 35.; Q. Tang, N. Wu, M. L. Lee, J. Microcolumn Separations, 12 (2000)6.; R. Asiaie, X. Huang, D. Farnan, Cs. Horvath. J. Chromatogr. A, 806 (1998) 251). In addition, interconnection of silica particles surface modified by Al chelate compounds (S. Ueno, K. Muraoka, H. Yoshimatsu, A. Osaka, Y Miura, Journal-Ceramic Society Japan, 109 (2001) 210.) and microwave sintering of silica particles (A. Goldstein, R. Ruginets, Y. Geffen, J. of Mat. Sci. Letters, 16 (1997) 310) have been reported. The interstitial porosity of the above particle-sintered or interconnected columns, and hence the permeability of the columns obtained by this approach is less than or similar to those of the conventional packed columns. Therefore, the backpressures of the column are the same or higher than those of the conventional packed columns, and result in an inability to achieve high efficiency chromatographic separations at low backpressures and high flow rates.
In an attempt to overcome the combined problems of packed bed stability and high efficiency separations at low backpressures and high flow rates, several groups have reported the use of monolith materials in chromatographic separations. Monolith materials are characterized by a continuous, interconnected pore structure of large macropores, the size of which can be changed independent of the skeleton size without causing bed instability. The large macropores allow liquid to flow directly through with very little resistance resulting in very low backpressures, even at high flow rates.
However there are several critical drawbacks associated with existing monolith materials. Columns made using organic monolith materials, e.g., polydivinylbenzene generally have low efficiency, particularly for low molecular weight analytes. Although organic monoliths are chemically stable against strongly alkaline and strongly acidic mobile phases, they are limited in the composition of organic solvent in the mobile phase due to shrinking or swelling of the organic polymer, which can negatively affect the performance of these monolithic columns. For example, as a result of monolith shrinking, the monolith can lose contact with the wall and thus allow the eluent to by-pass the bed, whereupon chromatographic resolution is dramatically decreased. Despite the fact that organic polymeric monoliths of many different compositions and processes have been explored, no solutions have been found to these problems.
In addition, chromatographic columns have also been made from inorganic monolith materials, e.g., silica. Inorganic silica monoliths do not show evidence of shrinking and swelling, and exhibit higher efficiencies than their organic polymeric counterparts in chromatographic separations. However, silica monoliths suffer from the same major disadvantages described previously for silica particles: residual silanol groups after surface derivatization create problems that include increased retention, excessive tailing, irreversible adsorption of some analytes, and the dissolution of silica at alkaline pH values. In fact, as the variation of the pH is one of the most powerful tools in the manipulation of chromatographic selectivity, there is a need to expand the use of chromatographic separations into the alkaline pH range for monolith materials, without sacrificing analyte efficiency, retention and capacity.
Hybrid silica monoliths offer a potential solution to overcome the problems of residual silanol group activity and hydrolytic instability of silica-based monoliths, wherein an alkyl functionality is coupled into both the skeleton structure and the surface of the silica. Several approaches are disclosed in U.S. Pat. No. 6,207,098 and Japanese patent application 2,893,104. However, the materials produced by the processes disclosed in these patents require a calcination step to form the porous gel. At temperatures above 400° C., many organic groups on the surface can be destroyed, leaving the surface unprotected. Furthermore, silanol groups can be irreversibly condensed above 400° C., leaving behind more acidic silanols. As a result, some analytes, particularly basic analytes, can suffer from increased retention, excessive tailing and irreversible adsorption. In addition, the hybrid monoliths produced by this process contain numerous micropores, i.e., pores having a diameter below 34 Å, which are known to inhibit solute mass transfer, resulting in poor peak shape and band broadening.
In an alternative approach, U.S. Pat. No. 5,869,152 discloses a hybrid silica wherein at least a portion of silicon is bonded to an alkyl moiety. However, the materials produced by this process are essentially unimodal in their pore size and lack mesopores to create a surface area sufficient for long analyte retention and high analyte loading capacity in a reversed-phase (RP) HPLC mode.